How do narratives shape economic reality?

The short answer

The intuition that markets are driven only by fundamentals — earnings, GDP, interest rates — has long been challenged by behavioral economics. Robert Shiller’s 2017 American Economic Association presidential address, expanded in his 2019 book “Narrative Economics,” took the argument further: the stories people tell about the economy spread through populations like epidemics and have measurable effects on consumption, investment and asset prices.

The intuition behind narrative economics is that humans process the world through stories, not through statistical models. A coherent story about why a stock will rise, why housing is safe, or why “this time is different” can shape behaviour across millions of decision-makers in ways that pure-fundamental models miss.

The complication is the reverse-causality problem. If asset prices rise first and narratives form afterward to explain the rise, the causation runs from reality to story, not from story to reality. Distinguishing the two empirically is hard, and Shiller himself acknowledges that the field is still developing methodologies for the test.

→ New to behavioral economics? Financial education hub

What the data shows

The empirical record on narratives draws on text-based research (Google Ngrams, news archives, social media) combined with traditional economic data.

The figures (Shiller, Google, news archives, 1929-2023):

• Shiller documents that the term “Great Depression” appeared in print frequencies that peaked in the early 1930s, returned in the 1970s, and surged again in 2008-2009 — each appearance correlated with worsening sentiment and consumer behavior
• Bitcoin’s narrative virality: media mentions of “Bitcoin” peaked near each major price peak (late 2013, late 2017, late 2021), generally lagging or coinciding with price moves rather than systematically leading them
• The 2017-2021 “FOMO” narrative around tech stocks coincided with NASDAQ doubling — but the price move began before the narrative reached saturation in mainstream press
• The “this time is different” narrative documented by Reinhart-Rogoff (2009) appears in nearly every major financial bubble, suggesting narrative pattern-repetition even when specific stories vary

The methodological challenge is that mass media coverage and price moves often co-evolve. Disentangling whether a narrative drives prices or prices drive narrative coverage requires either natural experiments (Shiller offers a few) or high-frequency identification, both of which are limited.

The exception worth flagging: in some cases the narrative evidence is unambiguous. Bank runs are by definition narrative-driven — depositors act on stories about bank solvency, and the act of acting on the story can make the story true. Silicon Valley Bank’s failure in March 2023 saw approximately $42bn in withdrawals in 24 hours after a narrative cascade through professional networks — the narrative was the proximate cause, not the consequence, of the run.

→ Dataset: CAPE ratio dataset

Why it happens — the macro mechanism

Three channels explain how narratives interact with economic outcomes.

Channel 1 — Coordination of expectations. A narrative provides a focal point for coordination. If most market participants believe “central banks will support markets at any sign of trouble,” each individual investor has rational reasons to underweight tail risk because others will support prices. The narrative becomes self-reinforcing not through delusion but through coordinated rational behavior.

Channel 2 — Demand and credit feedback loops. This is the angle most overlooked. Narratives that change consumer or investor demand feed through to financial-system credit creation. The 2003-2006 narrative of “housing prices never fall nationally” shifted mortgage underwriting standards (relaxed); relaxed underwriting expanded credit; expanded credit pushed housing prices higher; higher prices reinforced the narrative. The loop ran for several years before reversing in 2007-2008.

Channel 3 — Reverse causality. A meaningful share of narratives form after price moves, not before. The “AI boom” narrative of 2023-2024 accelerated after NVIDIA’s stock had already tripled — many investors were rationalizing existing positions, not creating new ones. Distinguishing leading from lagging narratives requires high-frequency text data and careful timing, and the conclusion in many cases is that narratives lag.

Synthesis by regime: in stable regimes, narratives are mostly explanatory and follow data; in transition regimes (build-ups to crises, late-stage bubbles), narratives become co-causal with prices through coordination and credit channels; in crisis regimes (acute panics, bank runs), narratives can dominate fundamentals over short windows because liquidity and solvency depend on coordinated belief. Three regimes, three causal weights.

Narratives are not separable causes of economic outcomes — they are part of the same dynamic, sometimes leading, sometimes lagging, and most powerful when fundamentals and stories reinforce each other.

→ Framework: Equity markets pillar

What it means for different economic actors

Active investors face the practical implication that narrative-driven moves can persist longer than fundamentals justify, and that fading narratives early (“shorting bubbles”) has historically been a costly strategy. Keynes’ observation that “markets can stay irrational longer than you can stay solvent” remains empirically defensible.

Passive investors have less to do with narratives operationally — their exposure tracks indices regardless. The narrative dimension affects them through aggregate market valuations, which is why CAPE-based measures and other valuation indicators retain relevance even for those who do not actively trade.

Policymakers face the most demanding interaction. A central bank that signals concern about a narrative (“irrational exuberance,” Greenspan 1996) may amplify the narrative simply by acknowledging it. Communication strategies have to account for the narrative-coordination effect.

A common error is to treat narratives as either everything (markets are pure storytelling) or nothing (markets are pure fundamentals). The empirical record supports a middle position: narratives matter at the margin and matter most in transitions, while fundamentals dominate over multi-year horizons.

Practical observation

Go deeper

Related questions

• Why is financial history the best teacher?
• Why do bubbles feel rational at the time?
• Why do economists disagree on basic questions?
• What is loss aversion in investment decisions?

Frequently asked questions

Are narratives more powerful in modern markets than historically?

The infrastructure for narrative diffusion (social media, financial Twitter, retail trading apps) is undeniably faster than in previous eras. Whether this translates to stronger narrative effects on prices is debated empirically. Some studies find evidence of faster transmission and larger amplitude in retail-heavy episodes (GameStop 2021); other studies find that institutional positioning still dominates pricing on most days even when retail narratives are intense. The infrastructure has changed; the relative weight of narrative versus fundamentals is less clearly resolved.

Can narratives be measured rigorously?

The methodologies have improved. Topic modelling on news archives, sentiment analysis on social media, and Google Trends data give quantifiable proxies for narrative diffusion. The remaining challenge is causal identification — showing that the narrative measurement leads price changes rather than co-evolves with them. Some studies use natural experiments (event studies around unexpected news) to isolate narrative effects; the broader empirical claim remains methodologically harder than typical macro identification problems.

Should investors try to ride narratives?

Empirically, momentum strategies — including those that ride narrative-driven trends — have produced positive returns over long samples in many asset classes. The risk is that narrative-driven moves end abruptly when the narrative is challenged, and timing the exit is difficult. Strategies that ride narratives with strict risk controls (trend-following with stop-losses) have a documented track record; strategies that ride narratives without controls have an ugly historical tail.

Last updated — 30 July 2026